Could not register the application: Timeout was reached (part 3)

[Aaron Paden <aaronbpaden gmail com>]

So awhile back I opened a bug[1] because nautilus wouldn't start.

Actually, I later found that it **would** start, but only after some
unreasonable amount time. Matthias Clasen suggests that this is
because of a broken session. He very well may be right — though I
don't know much of anything about sessions, so I have no idea how it
might have gotten broken. Mostly I just do "systemctl enable gdm" and
that's that. I do know that at one point I deleted all the config
files in $HOME, which temporarily fixed the problem, but it eventually
came back.

Recently I bought a used desktop with BIOS instead of EFI, so I
installed syslinux over gummiboot. That and the usual updates are the
only changes I've made recently, but I'm getting new symptoms that I'm
hoping give someone some clues. Now, I can't start nautilus even from
the terminal. I get these errors:

(nautilus:969): GLib-GIO-CRITICAL **:
g_dbus_interface_skeleton_unexport: assertion
'interface_->priv->connections != NULL' failed

(nautilus:969): GLib-GIO-CRITICAL **:
g_dbus_interface_skeleton_unexport: assertion
'interface_->priv->connections != NULL' failed
Could not register the application: Timeout was reached

(nautilus:969): Gtk-CRITICAL **: gtk_icon_theme_get_for_screen:
assertion 'GDK_IS_SCREEN (screen)' failed

(nautilus:969): GLib-GObject-WARNING **: invalid (NULL) pointer instance

(nautilus:969): GLib-GObject-CRITICAL **: g_signal_connect_object:
assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

I'm posting this on the list because it's been so long and the
symptoms have changed. I'm not even sure it's the same bug. It might
not even be a bug, and I might have broken my session in two or three
completely different ways!

Also, as you can imagine this is a pretty major inconvenience. Either
my session is broken (and I am inclined to believe the GNOME
developer), or Nautilus has a bug. If the first, what might have
caused it and how to I fix it/prevent it from happening again? If the
second, what information might contribute to a fix?

1: https://bugzilla.gnome.org/show_bug.cgi?id=727622